Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    INDEX
    Explanations

    contractions of "will" and terms related to future action or potential scenarios

    oai_token-act-pair · gpt-3.5-turbo

    modal verbs indicating future or conditional actions

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     bidding
    -0.87
     praying
    -0.81
     inspecting
    -0.77
     plotting
    -0.75
     searching
    -0.75
     burying
    -0.75
     conducting
    -0.74
     resisting
    -0.74
     manipulating
    -0.74
     performing
    -0.73
    POSITIVE LOGITS
     be
    1.19
     occur
    1.07
     coincide
    1.04
     suffice
    0.99
     prevail
    0.97
     arise
    0.97
     unfold
    0.95
     commence
    0.95
     remain
    0.93
     come
    0.92
    Act Density 0.269%

    No Known Activations